基于Apriori算法的机动车保险欺诈索赔的关联度分析  

Correlation Analysis of Motor Vehicle Insurance Fraud Claims Based on Apriori Algorithm

作　　者：金慧 杨涵予 崔浩浩 Jin Hui;Yang Hanyu;Cui Haohao

机构地区：[1]兰州工业学院汽车工程学院,甘肃省兰州市730050

出　　处：《时代汽车》2023年第9期180-183,共4页Auto Time

基　　金：甘肃省教育厅创新能力提升项目(2021B-315)。

摘　　要：在保险行业中,保险欺诈是当前存在的普遍现象。然而,在保险的全部险种里面,机动车车险欺诈是保险欺诈的高发区之一。本文基于25项特征指标,应用车险欺诈关联分析模型对某保险公司欺诈识别系统数据进行了分析,得出黑色样本的频繁项集,并且在白色样本中进行验证,确定了它们之间的具体关联规则,识别出欺诈索赔事件,得到灰色样本的欺诈率。研究结果表明,总理赔样本的欺诈率为37.527%。Insurance fraud is a common phenomenon in the insurance industry.However,among all types of insurance,motor vehicle insurance fraud is one of the most common areas for insurance fraud.Based on 25 characteristic indicators,this paper analyzes the data of the fraud identification system of an insurance company by applying the auto insurance fraud correlation analysis model,obtains the frequent itemset of the black sample,verifies it in the white sample,determines the specific association rules between them,identifies the fraud claim events,and obtains the fraud rate of the gray sample.The results of the study showed that the fraud rate of the Prime Minister's compensation sample was 37.527%.

关 键 词：车险欺诈 关联分析 APRIORI算法 欺诈率 

分 类 号：D92[政治法律—法学]